All Collections
Find your Insights
How do I filter my data by Sentiment?
How do I filter my data by Sentiment?

Learn how to use the Sentiment filter to isolate different sentiments expressed in customer feedback data.

Amanda Robinson avatar
Written by Amanda Robinson
Updated over a week ago

Keatext's different filters are essential tools for discovering actionable insights, identifying trends and diving deeper into the data. 

Sentiment Filters allow you to choose which sentiment categories you’d like to explore: Problems, Praises, Suggestions or Questions.

Classifying data into Sentiment categories

Keatext analyzes text by identifying relevant Topics in each sentence – the main subject of the idea expressed. Each Topic is then linked with an Opinion – how individuals qualified the service or products they are giving feedback on. Each Topic-Opinion pairing – including all its variations – becomes a Comment which is then classified into one of the four sentiment categories. 

Overview by Sentiment category

Once the data has been analyzed, you can see an overview at the top identifying the number of Praises, Problems, Suggestions and Questions.

Choosing the categories you wish to display

The sentiments act as filters. Each of these sentiment has a checkbox. For example, if you only want to see Problems, you can select the Problems checkbox and deselect the remaining options. 

If you want to see all the categories, click “Reset” in the upper left.

Use the adjustable Correlations chart in the Topics view to find more Sentiment insights

Volume trends reveal at what frequency different sentiments show up. Click on a Topic of your choice, move to the Correlations tab and choose the metadata parameter you wish to appear on the X-Axis by selecting from the dropdown using the Filters located at the top right of the chart.

Note: This can be done in the Comments view as well to reveal the volume by the chosen metadata parameter, however, only one Sentiment will appear as Comments are bucketed into one of the sentiments during the text mining phase.

Read more about Correlations in the article here.

Learn about these topics in Keatext’s Learning Hub:

Did this answer your question?